metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: completion
dtype: string
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
splits:
- name: train
num_bytes: 1893999
num_examples: 927
download_size: 763724
dataset_size: 1893999
configs:
- config_name: default
data_files:
- split: train
path: data/train-*